paperKB
coga / coga-kb
Processing
Help
Sign in

Chunk #15 — IRT models for polytomous data

Source
Variance decomposition using an IRT measurement model.
Embedded
yes

Text

Often, measurement is based on items or symptoms with more than two categories. For example, answers can be coded as 0 (not at all), 1 (somewhat, sometimes) and 2 (a lot, often). Typically in CTT approaches in behaviour genetics the sum of these item scores is regarded to represent a person’s score on the trait of interest and is used for the statistical inference.